Observing Brain Most Visited Common Band Connectivity States from fMRI Resting State Studies

Abstract

Neuroscientists have been working for years on finding the neural codes that can correlate neuron firing spatial and/or temporal patterns with behaviors to comprehend the mechanism of brain functions, predict behaviors, and identify methods to treat disorders. Due to the high spatial-temporal resolution requirement of such an approach, invasive measurement methods usually will be required. The other approach to study the mechanistic functions of brain spatial dynamics is using the activation statistics that are correlated to different types of tasks. Here we present rest state activation statistic results as baselines for later more advanced studies including our finding on “common bands” of these most visited brain connectivity states and the possible meaning of these findings. We bundle the MRI voxels to the thalamus (THL), basal ganglia (BSL), and 7 other cortical networks and use energy landscape analysis to explore connectivity signatures of them. Two different data sets obtained from two different fMRI tools were utilized. One dataset consists of 23 young adult and 47 old adult subjects with normal cognitive function. The other data set contains 107 schizophrenic patients and 86 healthy controls. We found that there are common bands of connectivity states that have consistently low energies in all 4 different groups of subjects. These brain-most visited states inside these bands are one or two hamming distances away from each other and centered around the BSL-THL core and then extended to the control type of cortical brain networks as well as other sensory networks.